Now that there’s no more campaigning to do locally, I’ll indulge in a little data-scrounging and even optimism. (I’ll write up the extraordinary turnout successes of our intense local effort in Champaign County, IL later.) Here’s a note I sent the NYT:
I think that it looks like CA-21 should be uncalled and returned to the bin of races to follow. Here's why.
On election night, R was up 5.2k, considered enough for a call.
Since then, ~19K votes have been added and now it's R up 2k.
Based on the 9k+ uncounted ballots for Kings county (entirely in CA-21, I think), extrapolated to the entire CD (the other counties are only partly in CA-21) there should be roughly 25k more ballots coming in. That's consistent with CA-21 ending up with about the same % of its net 2016 ballot as other CA districts.
If the remaining late-count ballots run nearly as D as the first late-count ballots, CA-21 will end up D.
This unusually heavy swing would not be a complete fluke, since CA-21 is an extreme for
- poverty
- youth
- minority
which all seem likely to increase this late-vote swing, consistent with the data that have come in so far.
Perhaps I've missed something, but this seems like the most likely prediction based on the standard public numbers.